This is JavaMI v1.1, an implementation of MIToolbox in Java.

It provides a series of functions for working with information theory. It also 
contains some variable manipulation functions to preprocess discrete/categorical
variables to generate information theoretic values from the variables. 

These functions are targeted for use with feature selection algorithms rather 
than communication channels and so expect all the data to be available before 
execution and sample their own probability distributions from the data.

Functions contained:
 - Entropy
 - Conditional Entropy
 - Mutual Information
 - Conditional Mutual Information
 - generating a joint variable
 - generating a probability distribution from a discrete random variable

The Java source files are licensed under the LGPL v3. 

Update History
20/09/2016 - v1.1 - Fixing an indexing bug. Migrating to maven for builds.
20/01/2012 - v1.0 - Initial Release